#INITIAL WRANGLING:

#Read in data and update to lower case:
willamette_fish_passage <- read_csv(here("data", "willamette_fish_passage.csv"))%>% 
  clean_names()

#Mutate date to class "Date" and convert data to time series data frame (tsibble):
willamette_ts <- willamette_fish_passage %>% 
  mutate(date = mdy(date)) %>%
  as_tsibble(key = NULL, index = date)

Willamette Falls in June 2012. | Credit: Peggy Sigler

Overview

Here we look at the seasonal variation in passage of Steelhead trout, Coho salmon, and Jack-Coho salmon through the Willamette Falls fish ladder in Williamette, Oregon. Our analysis displays the varying seasonal trends for each of these three species as well as changes over time, looking at data from 2001-2010. This provides important insights into the population dynamics of all three species, which are of great ecological, cultural and economic value.

Columbia River Basin salmon run map | Credit: www.bpa.gov

Data source: Columbia Basin Research, School of Aquatic & Fishery Sciences, University of Washington; additional data courtesy of U.S. Army Corps of Engineers, NWD and Chelan, Douglas, and Grant County PUDs, Yakima Klickitat Fisheries Project, Colville Tribes Fish & Wildlife (OBMEP), Oregon Department of Fish & Wildlife, Washington Department of Fish & Wildlife (2021). [http://www.cbr.washington.edu/dart/query/adult_graph_text]

Fish passage at the Willamette Falls, Willamette River (Oregon)

Tab 1: Time Series of fish passage

# Read in data
fish_data <- read_csv(here("data", "willamette_fish_passage.csv")) %>% 
  clean_names()

# Select for coho, jack coho and steelhead
fish_clean <- fish_data %>% 
  select("date", "steelhead", "coho", "jack_coho") %>% 
  mutate(across(where(is.numeric), ~ ifelse(is.na(.), 0, .))) %>% 
  mutate(date = mdy(date)) %>% 
  mutate(year = year(date))
# Wrangle data in preparation for time series
fish_ts <- fish_clean %>% 
  mutate(yearmonth = yearmonth(date)) %>% 
  as_tibble(key = NULL, index = yearmonth) 
# Steelhead salmon
ggplot(data = fish_ts, aes(x = date, y = steelhead, color = steelhead)) +
  geom_line(color = "palegreen4") +
  theme_minimal() +
  labs(x = "Year", y = "Steelhead Salmon", title = "Steelhead Trout Passage over Time")
**Figure 1:** This graph shows that trend in steelhead salmon migration through the Columbia River and gives projections to 2031.

Figure 1: This graph shows that trend in steelhead salmon migration through the Columbia River and gives projections to 2031. 

# Coho slamon
ggplot(data = fish_ts, aes(x = date, y = coho, color = coho)) +
  geom_line(color = "red3") +
  theme_minimal() +
  labs(x = "Year", y = "Coho Salmon", title = "Coho Salmon Passage over Time")
**Figure 2:** This graph shows that trend in coho salmon migration through the Columbia River and gives projections to 2031.

Figure 2: This graph shows that trend in coho salmon migration through the Columbia River and gives projections to 2031. 

# Jack Coho salmon
ggplot(data = fish_ts, aes(x = date, y = jack_coho, color = jack_coho)) +
  geom_line(color = "orange1") +
  theme_minimal()+
  labs(x = "Year", y = "Jack Coho Salmon", title = "Jack Coho Salmon Passage over Time")
**Figure 3:** This graph shows that trend in jack coho salmon migration through the Columbia River and gives projections to 2031.

Figure 3: This graph shows that trend in jack coho salmon migration through the Columbia River and gives projections to 2031.

From the times series graphs on steelhead trout, coho and jack coho salmon, we can see that:

  1. All three species have strong, seasonal migratory patterns through the Columbia River.

  2. Steelhead trout numbers have remained relatively constant and are predicted to remain constant through 2030 while coho and jack coho salmon numbers are expected to increase.

  3. Steelhead trout spend a greater time on average in the Columbia River than do either of the other two species.

Tab 2: Season Plots

willamette_long <- willamette_ts %>% 
  pivot_longer(cols = chinook_run:pink, 
               names_to = "species", 
               values_to = "counts") %>% 
  filter(species %in% c("coho", "jack_coho", "steelhead"))


coho_season <- willamette_long %>% 
  filter(species == "coho") %>% 
  gg_season(y = counts, 
            period = "year", 
            show.legend = FALSE)+
  ylim(0,1000)+
  theme_minimal()+
  labs(x= "", y = "", title = "Coho Salmon")

#coho_season

jack_coho_season <- willamette_long %>% 
  filter(species == "jack_coho") %>% 
  gg_season(y = counts, 
            period = "year")+
  theme_minimal()+
  ylim(0,1000)+
  labs(x= "", 
       y = "Number of Observations", title = "Jack-Coho Salmon")

#jack_coho_season


steelhead_season <- willamette_long %>% 
  filter(species == "steelhead") %>% 
  gg_season(y = counts, 
            period = "year", 
            show.legend = FALSE)+
  theme_minimal()+
  ylim(0,1000)+
  labs(x= "Month", y = "", title = "Steelhead Trout") 

#steelhead_season

# compound_season <- willamette_long %>% 
#   gg_season(y = counts, 
#             period = "year")+
#   facet_wrap(~species)+
#   theme_minimal()
# 
# compound_season

patch_season <- (coho_season/jack_coho_season/steelhead_season)+
  plot_annotation(title = "Coho, Jack-Coho, and Steelhead Seasonal Abundance in Willamette Falls \nfrom 2001-2010")

patch_season

Figure 4. Abundance of Coho, Jack-Coho and Steelhead seasonally observed in at Willamette Falls fish ladder Willamette, Oregon. These observations range from 2001-2010.Year is indicated by color. Data: Columbia River DART (Data Access in Real Time).

  • Coho and Jack-Coho are consistently at peak abundance in October, with at least small numbers of individuals present from August to November.
  • The number of Coho observed seems to have increased in recent years.
  • Observations of Steelhead are more diffuse throughout the year with the most observations ranging from February to July.

Tab 3: Summary statistics and analysis

willamette_fish_annual <- willamette_fish_passage %>%
  mutate(date = mdy(date)) %>%
  mutate(date_year = year(date)) 

coho_annual <- willamette_fish_annual %>%
  count(date_year, wt = coho)
jack_coho_annual <- willamette_fish_annual %>%
  count(date_year, wt = jack_coho)
steelhead_annual <- willamette_fish_annual %>%
  count(date_year, wt = steelhead)

 

ggplot()+ 
  geom_line(data = coho_annual, aes(x = date_year, y = n, color = "Coho"), size = 0.7) +
  geom_line(data = jack_coho_annual, aes(x = date_year, y = n, color = "Jack Coho"), size = 0.7) +
  geom_line(data = steelhead_annual, aes(x = date_year, y = n, color = "Steelhead"), size = 0.7) +
  scale_x_continuous(breaks=c(2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010), limits = c(2001,2010)) + 
  scale_color_manual(values = c( "Coho" = "red3", "Jack Coho" = "orange1", "Steelhead" = "palegreen4"))+
  labs(color = "Species")+
  scale_y_continuous(labels = scales::label_number_si())+
  labs(x = "Year",
       y = "Counts",
       title = "Annual counts by Species, Willamette Falls (2001-2010)") +
  theme_minimal()

Figure 5. Annual counts by species at the Willamette Falls fish ladder, Willamette River, Oregon (2001-2010). Red, orange, and green lines indicate counts for the species coho, jack coho, and steelhead, respectively. Data: Columbia River DART (Data Access in Real Time).

Summary

Annual counts of coho, jack coho, and steelhead species on the Willamette Falls fish ladder (2001-2010) show the following major trends:

● An overall low number of the jack coho species for the entire time period 2001-2010

● A marked increase in the counts of individuals of the coho species for the years 2009-2010

● A high number of individuals of the steelhead species for the period 2001-2004 followed by a relative decrease in the years 2005-2009 and a recent increase in the year 2010